The Belfast Social Malaise Project

The Belfast Social Malaise project (1971-73), funded by the Northern Ireland Community Relations Commission and based in Queen's University Belfast, was probably the first to map multiple social deprivation in any part of Ireland. Similar studies subsequently became relatively common at a national level in Ireland in the 1990s. However, whereas these more recent studies utilised small area census data to construct indicators of multiple deprivation, the Belfast study collected information on 20,563 individual cases of social need. Family breakdown, health, crime, unemployment and housing were identified as key problem themes. Each individual was assigned to one of 97 specially defined social malaise project zones. They were also assigned to a 200 metre grid square. The total numbers of cases of each type of problem within each social area and each grid square were counted, divided by measures of the corresponding population at risk, and the resulting rates were then combined using principal components analysis and factor analysis to create composite social malaise indicators. These composite measures were mapped to highlight the areas with the greatest concentrations of social deprivation - mostly in the central and western parts of the city.

Similar studies had been conducted in the previous couple of years by the Local Authorities in Liverpool, Leeds and Glasgow. These studies had access to the full resources of those cities' planning departments, whilst the constituent indices were provided by the government agencies etc. responsible for the data. The Belfast study, in contrast, was conducted by two research assistants (Paul Doherty and myself) under the direction of Dr. Fred Boal, and all the data had to be manually transcribed, albeit with some additional assistance, from a variety of primary data sources. The Belfast study also departed from its predecessors by extending the study area beyond the administrative boundaries of the city to include the entire built-up area - an area which, in the case of Belfast, included about 40 per cent of the entire population of Northern Ireland.

Data were collected on a range of variables for each individual. Whilst this was very time consuming, it provided a rich body of information which added considerable depth to the analysis based on the simple and composite indicators. However, to safeguard the confidentiality of each individual, no details were recorded on each person's name or address. Instead, each individual was designated two geocodes (a zone number and a 200 metre grid square reference) indicating their place of normal residence. To facilitate this, we had to construct a code book (similar to the Geodirectory created by An Post and Ordnance Survey Ireland in the late 1990s) listing the codes for every single house in the Belfast Urban Area before we could even begin extracting the source data. This alone took several months to complete.

Once the data had been collected, it had to be transcribed onto punched cards for computer analysis. It then had to be carefully checked for errors and corrected - another painstaking and lengthy process. Finally the data had to be analysed and mapped.

No suitable software existed at that time, so we had to write our own. This software enabled us to count the number of cases in each zone or grid square, merge the resulting counts with data on the population at risk to create rates (as well as calculate various other types of indicator - e.g. means), export the results on punched cards for principal component or factor analysis using an early version of SPSS (Statistical Package for the Social Sciences), and then re-import the results back into the system for mapping. This software, referred to as the Queen's University Social Monitoring System (QSMS), had many of the characteristics of a modern GIS and was almost certainly the first computer mapping program either used or developed in Ireland. (See the QSMS page for further details). It was hoped that the QSMS would enable the study to be readily and rapidly replicated, and that a social monitoring system which could advise policy makers might be established, but unfortunately this was not to be.

The Belfast Social Malaise study was purely descriptive. No attempt was made to explore the underlying causes of the problems, although suggestions were made as to the directions that such an exploration might usefully take. However, the multivariate statistical techniques allowed the indicators to be examined using three different models of their interrelationships, resulting in 3 definitions of social malaise.

The first definition considered social malaise to be the incidence of a number of equal problems in an area. A ranking method of analysis was employed, which revealed a central city concentration of malaise, with a western sectoral bias.

The second definition of social malaise gave weight to problems which have high spatial intercorrelations. An unrotated principal components analysis was used to concentrate the maximum variance onto the first component, the spatial distribution of which was again focussed on the inner city and the western sector.

The third definition considered social malaise to be a set of underlying social processes, of which the individual problem indicators are merely symptoms. A classical factor analysis with oblique rotation was employed. The first factor extracted was quite general in nature, with an inner city distribution. The second factor was characterised by crowding and unemployment, and had a central and western sectoral distribution. The relationships of definition 3 social malaise to a set of general social characteristics was examined in a later (1974) paper published by Queen Mary College London to help in understanding the underlying social processes.

Return to the main Social Geography page
Return to the Doctorate Thesis page
Return to the Programming Background page

[Home Page] [Computers/GIS] [Medical] [Political] [Social]
[Index Of Other DGP Sites]